Scen-O-Con

SDP matrices are entered as a collection \(F_0, F_1, \ldots, F_d\) (or \(E_0, \ldots, E_d\)). Set \(n\) = number of decision variables and \(m\) = matrix dimension, then click Select to create buttons for each matrix.

File upload (all at once):

A collection requires a dict-of-matrices structure, so only JSON and MAT are supported here.

• JSON: Dictionary keyed by index ("0", "1", ...), each value a 2D matrix.

  {"0": [["delta[0]", "0"], ["0", "1"]],
   "1": [["1", "0"], ["0", "-1"]], ...}

• MAT: MATLAB .mat file where each variable is one matrix in the collection.

  % MATLAB: save('F_d.mat', 'F0', 'F1', 'F2', 'F3')
  % Variable names become dict keys

Individual entry:

Click each \(F_i\) / \(E_i\) button to open a matrix editor for that index. Individual matrices accept all formats (CSV, JSON, TXT, TSV, MAT, Excel, NPY, Parquet) and can contain numbers or expressions with delta[i].

Matrices can contain numeric values or expressions using the delta variable (e.g., delta[0] + 1, sin(delta[1])).

Supported file formats:

• CSV/TXT: Comma-separated values, one row per line.

  delta[0]+1, 0
  0, delta[1]

• TSV: Same as CSV but tab-separated.
• JSON: 2D array of strings or numbers.

  [["delta[0]+1", "0"], ["0", "delta[1]"]]

• MAT: MATLAB .mat file. The first non-metadata variable is used as the matrix.

  % MATLAB: save('A_d.mat', 'A_d')
  % where A_d = [1 0; 0 -1]

• Excel: .xlsx / .xls file. Reads the first sheet with no header row. Each cell becomes a matrix entry.
• NPY: NumPy .npy file containing a 2D array.

  # Python: np.save('A_d.npy', A_d)

• NPZ: NumPy .npz archive. The first array in the archive is used.
• Parquet: Apache Parquet file. Columns become matrix columns; rows become matrix rows.

Manual entry:

Set the dimensions with Rows/Columns and click Resize, then fill in each cell. Cells accept numbers or expressions with delta[i].

The file must include a type field (LP, QP, or SDP) along with the required matrices and parameters.

Required fields by type:

• LP: type, A_d, b_d, c. Optional: G, h, rho, tau, confidence, p, x_ref.
• QP: Same as LP, plus Q.
• SDP: type, F_d (dict of expression matrices keyed "0","1",...), c. Optional: E (dict of numeric matrices), Q, rho, tau, confidence.

JSON format:

{
  "type": "LP",
  "A_d": [["-delta[0]", "0"], ["0", "-delta[0]"]],
  "b_d": [["delta[0] - 1"], ["delta[0] - 1"]],
  "G": [["-1", "0"], ["0", "-1"]],
  "h": [["0"], ["0"]],
  "c": [["1"], ["1"]],
  "rho": "0",
  "tau": "0",
  "confidence": "0.01"
}

MAT format:

Each MATLAB variable maps to a field name. Variable names must match the required fields above.

% MATLAB example for LP:
type = 'LP';
A_d = [-1 0; 0 -1];  % numeric matrices
b_d = [1; 1];
c = [1; 1];
rho = 0; tau = 0; confidence = 0.01;
save('benchmark.mat', 'type', 'A_d', 'b_d', 'c', ...
     'rho', 'tau', 'confidence')